Share this article on:

Effect of Lactobacillus GG and Breast-feeding in the Prevention of Rotavirus Nosocomial Infection

Mastretta, Emmanuele*; Longo, Patrizia*; Laccisaglia, Anna*; Balbo, Luciano; Russo, Roberto; Mazzaccara, Alfonso; Gianino, Paola*

Journal of Pediatric Gastroenterology and Nutrition: October 2002 - Volume 35 - Issue 4 - p 527-531
Original Articles: Gastroenterology

Background Rotavirus is one of the leading etiologic agents of nosocomial infections among children. The development of preventive measures is therefore important. The efficacy of Lactobacillus GG in the treatment of rotavirus infection has been reported in literature, but there is only one recent study about its effectiveness in prevention of infection. The role of breast-feeding in the prevention of rotavirus infection is still debated. The aim of our study was to assess the efficacy of Lactobacillus GG and breast-feeding in the prevention of nosocomial rotavirus infections.

Methods In a randomized, placebo-controlled, double-blind study, 220 children aged 1 to 18 months hospitalized from December 1999 to May 2000, received Lactobacillus GG (n = 114) at a dose of 1010 colony-forming units or a comparable placebo (n = 106) every day of their hospital stay. Rotavirus testing on stool samples was performed for every patient on admission, during hospitalization, and after discharge.

Results The total incidence of nosocomial rotavirus infections was 27.7% (61 of 220 patients). The attack rate of rotavirus infections among the patients who received probiotic was 25.4% (29 of 114 patients), while for the placebo group it was 30.2% (32 of 106 patients). The difference is not significant (P = 0.432). Forty-seven of 220 infants (21.4%) were breast-fed, and 173 of 220 (78.6%) were non–breast-fed. The attack rate of rotavirus infections among breast-fed infants was 10.6% (5 of 47 infants), while for non–breast-fed infants it was 32.4% (56 of 173 infants). The difference is significant (P = 0.003).

Conclusion In our study, Lactobacillus GG was ineffective in preventing nosocomial rotavirus infections, whereas breast-feeding was effective.

*Departments of Pediatrics and †Clinical Pathology, Regina Margherita Children's Hospital, University of Turin, and ‡Department of Public Health, University of Turin, Turin, Italy

Received February 14, 2001; accepted May 14, 2002.

Address correspondence and reprint requests to Dr. Paola Gianino, Dipartimento di Scienze Pediatriche e dell'Adolescenza, Ospedale Infantile Regina Margherita Piazza Polonia 94, 10126 Torino, Italy (e-mail:

Rotavirus is one of the most important etiologic agents of nosocomial infections in childhood (1–2). Attempts to prevent nosocomial shedding of the virus have been unsuccessful so far (3). Despite the adoption of norms regarding enteric isolation (4), the incidence of nosocomial rotavirus infections is still high, ranging from 8 to 33 cases per 100 hospitalized children (5–9), depending on the age range of patients and the season of the year. As the vaccine is not available in Italy, the development of other preventive measures is important (10).

Evidence shows Bifidobacterium bifidum and Streptococcus thermophilus are useful in the prevention of rotavirus nosocomial diarrhea, as reported by Saavedra et al. (11), and that Lactobacillus GG (LGG; American Type Culture no. 53103) is protective against traveler's diarrhea (12) as well as antibiotic-associated diarrhea (13). Moreover, LGG seems to be effective in the treatment of rotavirus infection, as it shortens the symptomatic period and time of fecal shedding of the virus (14–16). Only one recent study (17) investigated the role of LGG in the prevention of rotavirus diarrhea.

There are conflicting data concerning the role of breast-feeding as a protective factor against rotavirus infection. Some investigators (18–22) state that the overall rates of rotavirus infection are similar in breast-fed and non–breast-fed infants. Others (20–23) indicate that, although rotavirus infection occurs in breast-fed infants, the severity of symptoms is substantially less than in non–breast-fed infants. Only a few studies (24–26) conclude that there is a protective effect of breast-feeding, especially in the newborn (27,28).

In this study we evaluated the effectiveness of LGG and breast-feeding in the prevention of nosocomial rotavirus infections.

Back to Top | Article Outline


The study was conducted in children aged 1 to 18 months, admitted for common diseases to the infant section of the Department of Paediatrics, University of Turin, between December 1, 1999, and May 31, 2000. Informed written consent was obtained from the parents before enrollment of their children.

To investigate a homogeneous group of patients unaffected by concomitant pathologies that might have modified the results, we excluded children with a diagnosis of enteritis at the time of admission or onset of enteritis within 24 hours after admission (community-acquired infections), those readmitted to hospital within 72 hours from the previous discharge (for the impossibility to distinguish a nosocomial from a community origin of the infection), those with a history of gastroenteritis in the 2 weeks before hospitalization (to avoid cases of gut microflora alterations or occasional postenteritis shedding of the virus) (29), and those with a history of immunodeficiency. Children were also excluded if hospitalization was voluntarily interrupted by parents (for the impossibility to complete the follow-up), or if their hospitalization was less than 48 hours (this period is not adequate for gut colonization by probiotics) (30).

On admission, each patient was randomly assigned, using odd and even random sampling numbers (31), to receive two capsules of 1010 colony-forming unites of LGG on admission and one capsule of 1010 colony-forming units of LGG once daily for the duration of hospitalization (treated group), or a comparable placebo made of inert oligosaccharides with indistinguishable organoleptic properties from the probiotic (placebo group). Both LGG and the placebo were manufactured and supplied by Dicofarm SpA (Rome, Italy). The capsules of LGG and placebo were opened and dissolved in water and administered before meals.

A case of nosocomial diarrhea was defined as the occurrence of three or more loose stools per day at least 24 hours after admission.

Rotavirus was searched for on fecal specimens collected from each child on admission to identify the asymptomatic carriers of rotavirus, during hospitalization if children developed diarrhea, at the time of discharge, and 72 hours afterward in every patient who did not develop nosocomial diarrhea, as the incubation period of rotavirus is 24–72 hours. Rotavirus antigen was detected in stool samples by a qualitative enzyme immunoassay test (Pathfinder™ Kallestad, Austin, TX).

Parents were given a container at the time of discharge and were asked to collect a stool sample 3 days later. If children developed enteritis within 72 hours after discharge, the samples were to be collected at the onset of diarrhea. The container, stored in a refrigerator, was returned at the visit to hospital 4 days after discharge.

We completed a data record form for each patient, reporting sex, age, period of hospitalization, breast-feeding or other feeding, antibiotic administration, diarrhea onset during hospital stay or after discharge, and rotavirus detection on stool samples at the time of admission, during hospitalization, and after discharge. Neither the medical staff nor the parents knew to which group the patients had been randomly assigned.

Examination of fecal specimens and statistical investigation were conducted blindly. The χ2 and Fisher exact tests were used for comparisons of proportions. Relative risk (RR) and 95% confidence limits were calculated using Epi-info version 6 software (CDC, Atlanta, GA, USA). The differences between the groups were considered significant when the P value was < 0.05. The identification codes of probiotic and placebo were kept in a sealed envelope and were opened at the end of data processing.

Back to Top | Article Outline


During the study period, 420 patients were hospitalized. We excluded 126 patients with admission diagnoses of enteritis, 5 patients who developed enteritis within 24 hours of admission, 3 patients who were readmitted within 72 hours from the previous discharge, and 17 patients with a history of gastroenteritis in the 2 weeks before hospitalization. Of the remaining 269 patients, none was affected by immunodeficiency. The patients were randomly assigned to each group: 134 to the treated group and 135 to the placebo group.

Rotavirus detection on stool samples collected during the first day of hospitalization led to the exclusion of 20 of 269 patients (7.4%) who were asymptomatic carriers of the virus (8 belonged to the treated group and 12 to the placebo group). Moreover, 22 patients were excluded as they voluntarily interrupted their hospitalization, and 7 patients did not come to the visit after discharge. Of these 29 dropout patients, 12 belonged to the treated group and 17 to the placebo group.

Thus, the study was conducted in 220 patients, 114 belonging to the treated group and 106 to the placebo group (Table 1); 123 were male (55.9%) and 97 female (44.1%). There was not a significant sex-related difference between the groups (P = 0.173). The mean age was 10 months (median, 8 months; range, 1–18 months). The patients were homogeneously distributed between the groups in accordance with age range (≤ 6 months or > 6 months), preadmission antibiotic administration, and type of alimentation (breast-feeding or other feeding). The reasons for hospitalization of the study patients are shown in Table 2.





The incidence of nosocomial rotavirus infection (patients with negative rotavirus detection on the first sample taken on admission, who produced a positive test during hospitalization or within 72 hours after discharge) was 27.7% (61 of 220 patients;Table 3).



Of the 61 infected patients, 37 were symptomatic (60.7%) and 24 asymptomatic (39.3%). Thus, the incidence of nosocomial symptomatic rotavirus infection was 16.8%, and that of asymptomatic infection was 10.9%; the difference is not significant (P = 0.098). Of the 61 infections, 26 cases occurred during hospitalization (incidence rate = 11.8%), while 35 occurred after discharge (incidence rate = 15.9%); the difference is not significant (P = 0.268).

Nine cases had nosocomial nonrotavirus enteritis, which together with the 37 rotavirus symptomatic infections increased the incidence of nosocomial enteritis to 20.9% (46 of 220 patients). The attack rate of nosocomial rotavirus infection among the patients who received the probiotic was 25.4% (29 of 114 patients), whereas in the placebo group it was 30.2% (32 of 106 patients;Table 4); the lower risk assessed for the treated group (RR = 0.84) is not significant (P = 0.432).



We detected 15 cases (13.2%) of symptomatic infection in patients receiving LGG versus 22 cases (20.8%) in the placebo group. Even if we assessed a lower risk of symptomatic infection in the treated group (RR = 0.63), the difference is not significant (P = 0.132). The higher risk of asymptomatic infection in the treated group (RR = 1.30) is not significant either (P = 0.499).

Of the 26 cases with infection during hospitalization, 12 belonged to the treated group and 14 to the placebo group (RR = 0.80); of the 35 cases identified in the period after discharge, 17 belonged to the treated group and 18 to the placebo group (RR = 0.98). The differences are not significant.

The total number of days spent in hospital by the enrolled patients was 1,144. The mean duration of hospital stay in the treated group was 5.33 days, while in the placebo group it was 5.1 days. The difference is not significant.

Forty-seven of 220 infants (21.4%) were breast-fed (at least 2 meals a day), and 173 of 220 (78.6%) were non–breast-fed. Only 5 of 47 breast-fed infants (10.6%) developed nosocomial rotavirus infection, versus 56 of 173 non–breast-fed infants (32.4%) (Table 5); the difference is significant (P = 0.003). Of the five breast-fed infants who developed infection, two belonged to the treated group and three to the placebo group; the lower risk found for the treated group (RR = 0.49) is not significant (P = 0.404). Among the 56 non–breast-fed children with rotavirus infection, 27 were in the treated group and 29 in the placebo group. Also in this case, the lower risk assessed (RR = 0.92) is not significant (P = 0.706).



None of the 5 breast-fed children who contracted rotavirus infection had diarrhea, while 37 of 56 (66%) of the non–breast-fed children did. The difference is significant (P = 0.007).

Back to Top | Article Outline


Our study demonstrates a high incidence of nosocomial rotavirus infections (27.7%) for several reasons. The study was performed during the seasonal peak of rotavirus infections, from December to May, and because the age range studied was at the highest risk of contracting rotavirus infection. The data collected concerned both symptomatic and asymptomatic infections. Although the latter is less harmful for the individual patient, but are responsible for virus shedding, both in hospitals and in child communities (kindergarten, nursery schools, etc.) (32). Stool samples were collected not only during hospitalization, but also within 72 hours after discharge, allowing us to identify 35 of 61 nosocomial infections (57.4%).

The importance of asymptomatic rotavirus infections has been emphasized in reports concerning patients admitted for other pathologies (33), children attending infant communities (32), and also children without any risk factor (34). The importance is emphasized by our study, considering that 39.3% of the nosocomial infections were asymptomatic. The frequent turnover of patients in our unit (mean hospital stay = 5.2 days) and the incubation period of rotavirus (1–3 days) may help to explain the high rate of nosocomial infections identified after discharge.

Our data show that LGG is ineffective in the prevention of nosocomial rotavirus infections or in protecting against the onset of diarrhea symptoms in infected patients.

LGG is a physiologic constituent of gut microflora. It is unaffected by gastric acid (35) (it survives up to pH 3) and adheres to intestinal mucosa (36), a necessary condition for stable intestinal colonization. Colonization of the stool has been shown to be dosage-dependent and occurs in all patients if given at a dose of 1010 colony-forming units per day for at least 48 hours (30).

One source of error in this study could be the concomitant use of antibiotics. However, owing to randomization, antibiotic treated patients were homogeneously distributed into the treated group and the placebo group. Moreover, LGG is resistant to the most widely used antibiotics, as reported in the first studies by Goldin et al. (35), where LGG was recovered in the feces of 20 of 23 patients who were concurrently receiving ampicillin. Recently, LGG proved to be resistant to a group of antibiotics, including cefoxitin, aztreonam, amikacin, gentamicin, kanamycin, streptomycin, norfloxacin, nalidixic acid, sulphametoxazole, trimethoprim, cotrimoxazole, metronidazole, polymixin, and colistin sulphate (37). LGG was susceptible to tetracycline, chloramphenicol, and rifampicin, which were not used in the study patients. Finally, LGG has proved to be effective in the prevention of antibiotic-associated diarrhea (38).

A second bias could be the concomitant breast-feeding, but in this instance as well, breast fed children were evenly distributed between the groups. As the mechanisms of LGG therapy is unclear, it is difficult to explain why it has been proved to be effective in the prevention of antibiotic-associated diarrhea.

In a recent study (17), which was similar to ours for the age of patients, probiotic dosage, and treatment length, LGG did not prove to be effective in the prevention of rotavirus infections, but it did prevent diarrhea symptoms.

The efficacy of breast-feeding in the prevention of rotavirus infections is still under discussion in literature. In our study, breast-feeding was associated with a reduced incidence of diarrhea symptoms in the infected patients and a reduced incidence of rotavirus nosocomial infections. It has been recently reported that lactadherin, a glycoprotein found in different concentrations in human milk, may bind rotavirus and inhibits its replication, reducing the symptoms of the infection (39). We speculate that this protein is implicated in preventing infection in our breast fed patients.

In conclusion, our findings fail to support the efficacy of LGG in the prevention of rotavirus infections and in preventing diarrhea symptoms, whereas breast-feeding proved to be effective.

Back to Top | Article Outline


1. Haffejee IE. The epidemiology of rotavirus infections: A global perspective. J Pediatr Gastroenterol Nutr 1995; 20:275–86.
2. Berner R, Schumacher RF, Hameister S, Forster J. Occurrence and impact of community-acquired and nosocomial rotavirus infection: A hospital-based study over 10 years. Acta Paediatr Suppl 1999; 426:48–52.
3. Dennehy PH, Tente WE, Fischer DJ, Veloudis BA, Peter G. Lack of impact of rapid identification of rotavirus infected patients on nosocomial rotavirus infection. Pediatr Infect Dis J 1989; 8:290–6.
4. Garner JS. Guidelines for isolation precautions in hospitals. The Hospital Infection Control Practices Advisory Committee. Infect Control Hosp Epidemiol 1996; 17:53–80.
5. Ford-Jones EL, Mindorff CM, Gold R, Petric M. The incidence of viral-associated diarrhea after admission to a pediatric hospital. Am J Epidemiol 1990; 131:711–8.
6. Cone R, Mohan K, Thouless M, Corey L. Nosocomial transmission of rotavirus infection. Pediatr Infect Dis J 1998; 7:103–9.
7. Koopmans M, Van Asperen I. Epidemiology of rotavirus infections in The Netherlands. Acta Paediatr Suppl 1999; 426:31–7.
8. Dennehy PH, Peter G. Risk factors associated with nosocomial rotavirus infection. Am J Dis Child 1985; 139:935–9.
9. Tufvesson B, Johnsson T. Occurrence of reo-like calf viruses in young children with acute gastroenteritis: Diagnoses established by electron microscopy and complement fixation, using the reo-like virus as antigen. Acta Pathol Microbiol Scand [B] 1976; 84:22–8.
10. Kapikian AZ, Flores J, Hoshino Y, et al. Prospects for development of a rotavirus vaccine against rotavirus diarrhea in infants and young children. Rev Infect Dis 1989; 3(suppl):539–46.
11. Saavedra JM, Bauman NA, Oung I, Perman JA, Yolken RH. Feeding of Bifidobacterium bifidum and Streptococcus thermophilus to infants in hospital for prevention of diarrhea and shedding of rotavirus. Lancet 1994; 344:1046.
12. Oksanen PJ, Salminen S, Saxelin M, et al. Prevention of traveller's diarrhoea by Lactobacillus GG. Ann Med 1990; 22:53–6.
13. Siitonen S, Vapaatalo H, Salminen S, et al. Effect of Lactobacillus GG yoghurt in prevention of antibiotic associated diarrhoea. Ann Med 1990; 22:57–9.
14. Isolauri E, Juntunen M, Rautanen T, Sillanaukee P, Koivula T. A human Lactobacillus strain (Lactobacillus Casei strain GG) promotes recovery from acute diarrhea in children. Pediatrics 1991; 88:90–7.
15. Pant AR, Graham SM, Allen SJ, et al. Lactobacillus GG and acute diarrhoea in young children in the tropics. J Trop Pediatr 1996; 42:162–5.
16. Guarino A, Canani RB, Spagnuolo MI, Albano F, Di Benedetto L. Oral bacterial therapy reduces the duration of symptoms and of viral excretion in children with mild diarrhea. J Pediatr Gastroenterol Nutr 1997; 25:516–9.
17. Szajevska H, Kotowska M, Mrukovicz JZ, Armanska M, Mikolajezyk W. Efficacy of Lactobacillus GG in prevention of nosocomial diarrhea in infants. J Pediatr 2001; 138:361–5.
18. Weinberg RJ, Tipton G, Klish WJ, Brown MR. Effect of breast-feeding on morbidity in rotavirus gastroenteritis. Paediatrics 1984; 74:250–3.
19. Glass RI, Stoll BJ, Wyatt RG, Hoshino Y, Banu H, Kapikian AZ. Observation questioning a protective role for breast-feeding in severe Rotavirus diarrhea. Acta Paediatr Scand 1986; 75:713–8.
20. Duffy LC, Riepenhoff-Talty M, Byers TE, et al. Modulation of rotavirus enteritis during breast-feeding: Implications on alterations in the intestinal bacterial flora. Am J Dis Child 1986; 140:1164–8.
21. Clemens J, Rao M, Ahmed F, et al. Breast-feeding and the risk of life-threatening rotavirus diarrhea: Prevention or postponement? Pediatrics 1993; 92:680–5.
22. Gurwith M, Weirman W, Hinde D, et al. A prospective study of rotavirus infection in infants and young children. J Infect Dis 1981; 114:218–24.
23. Berger R, Hadziselimovic F, Just M, Reigel F. Influence of breast milk on nosocomial rotavirus infections in infants. Infection 1984; 12:171–4.
24. Howie PW, Forsyth JS, Ogston SA, Clark A, Florey CD. Protective effect of breast-feeding against infection. Br Med J 1990; 300:11–6.
25. Cunningham AS. Breast-feeding and health. J Pediatr 1987; 110:658–9.
26. Blake PA, Ramos S, Mac Donald KL, et al. Pathogen specific risk factors and protective factors for acute diarrheal disease in urban Brasilian infants. J Infect Dis 1993; 167:627–32.
27. Totterdell BM, Chrystie IL, Banatvala JE. Rotavirus infections in a maternity unit. Arch Dis Child 1976; 51:924–8.
28. Chrystie IL, Totterdell BM, Banatvala JE. Asymptomatic endemic rotavirus infections in the newborn. Lancet 1978; 1:1176–8.
29. Eiden JJ, Verleur DG, Vonderfecht SL, Yolken RH. Duration and pattern of asymptomatic rotavirus shedding by hospitalized children. Pediatr Infect Dis J 1988; 7:564–9.
30. Saxelin M, Elo S, Vapaatalo H. Dose response colonisation of faeces after oral administration of Lactobacillus casei strain GG. Microb Ecol Health Dis 1991; 4:209–14.
31. Armitage P, Berry G. Statistical methods in medical research. Oxford, Blackwell Scientific, 1987, pp 172–5.
32. Pickering LK, Bartlett AV, Reves RR, Morrow A. Asymptomatic excretion of rotavirus before and after rotavirus diarrhea in children in day care centers. J Pediatr 1988; 112:361–5.
33. Walther FJ, Bruggeman C, Daniëls-Bosman MSM, et al. Symptomatic and asymptomatic rotavirus infections in hospitalized children. Acta Paediatr Scand 1983; 72:659–63.
34. Champsaur H, Questiaux E, Prevot J, et al. Rotavirus carriage, asymptomatic infection, and disease in the first two years of life: I. Virus shedding. J Infect Dis 1984; 149:667–74.
35. Goldin BR, Gorbach SL, Saxelin M, Barakat S, Gualtieri L, Salminen S. Survival of Lactobacillus species (strain GG) in human gastrointestinal tract. Dig Dis Sci 1992; 37:121–8.
36. Chauviere G, Coconnier MH, Kerneis S, Fourniat J, Servin AL. Adhesion of human Lactobacillus acidophilus strain LB to human enterocyte-like Caco-2 cells. J Gen Microbiol 1992; 138:1689–96.
37. Charteris WP, Kelly PM, Morelli L, Collins JK. Antibiotic susceptibility of potentially probiotic Lactobacillus species. J Food Prot 1998; 61:1636–43.
38. Arvola T, Laiho K, Torkkeli S, et al. Prophylactic Lactobacillus GG reduces antibiotic-associated diarrhea in children with respiratory infections: A randomized study. Pediatrics 1999; 104:1121.
39. Newburg DS, Peterson JA, Ruiz-Palacios GM, et al. Role of human-milk lactadherin in protection against symptomatic rotavirus infection. Lancet 1998; 351:1160–4.

Children; Rotavirus; Cross-infection; Prevention; Breast-feeding; Lactobacillus GG

© 2002 Lippincott Williams & Wilkins, Inc.